Big Blue to take on Pixar?
spareacct1 writes "USAToday is reporting that IBM is set to announce a strategic partnership with Threshold Digital Research Labs of Santa Monica, CA. TDRL now hopes it has the deep pockets and computing power to take on Pixar as the undisputed leader in CG animated films. TDRL's spartan website is showing off digital stills. Interesting sidebar at the end of the story, both Pixar and TDRL recently dumped Sun and MS, respectively, in favor of Linux."
Pixar's movies are good because of their people, not their computers. They've got good artists, good directors, and amazing writers. Without those, you end up with movies like Final Fantasy: technically adept, but ultimately empty and pointless.
You don't just need big computing power, you need design & drawing skills, besides lots of creativity and imagination.
I could have all the computing power and still not be able to do something worth watching.
well, believe it or not, but PIXARs success isn't really because of their rendering power -- true enough, the realism and rendering techniques used in their latest productions has contributed to making images better, but they've always had the edge when it comes down to the thing that matters: storytelling and keeping the audience interested. Look at their older shorts and their more recent feature films, the story is the main driving force.
While Final Fantasy looked quite amazing, the story and the movie just didn't fit in like most of the PIXAR movies. PIXAR makes movies for the whole family which people enjoy on different levels (best example, toy story 2) -- Shrek was a very welcome break from the PIXAR dominance, but not because it wasn't made by pixar, more because of a great story supported by a nice screenplay and good animation (it's more about how you use the tools, not that the end result has been raytraced with molecular precision)..
If they're able to produce films that would be entertaining even if they were hand drawn by a five year old, then the rendering power comes to good use; not the other way around.
mats
One man's ceiling is another man's floor.
So IBM's supplying hardware as a showcase of their new initiative. It's hardly 'taking on Pixar'. I bet IBM would love to do business with Pixar, too. Do people say that IBM's "taking on the XBox" by supplying the processor in the Gamecube?
What did you eat today? http://www.atetoday.com/
Hire some talented writers and storytellers. Contrary to popular opinion, Pixar's success had much less to do with the CGI than with traditional old storytelling skill. Pumping money into the technology side at the expense of people will result in a big financial loss.
Just ask Sony.
-- You see, there would be these conclusions that you could jump to
Sure, processing power is a crucial part of it, but it's only necessary for advancing the state-of-the-art in computer graphics.
In each movie that Pixar takes, it takes about 8 hours to render each frame (or so I've read in numerous locations) and you can see that with the increasingly "less-computery" look of their movies as processing power has increased for each one.
This brings me to the point that I'm intending on making: the realism of the graphics is not what makes a great movie, it's the quality of the story and all that. I saw Toy Story again the other week and it looks so dated now compared to say Monsters Inc. It was still a thoroughly entertaining movie though because it was a good story.
I love CG films, but I admit that the main reason I love seeing them is to see what new effects and advancements have been made, which is why Pixar films are so great to me.. they're always advancing the state-of-the-art.
Damnit, now I've just contradicted the original point I was trying to make! Hrmm... BRING ON THE CG FILMS!
Regardless of whether the G5 is the fastest CPU for RenderMan, it is not per-CPU performance that matters. If you're setting up a rendering farm, you're buying n computers to render m frames per hour. At the end of the day, what matters is minimising $$$$$$ per m, not n, and I'll bet dollars to doughtnuts that commondity Intel/AMD whips a G5 mac in terms of rendered frames per dollar. Remember, Apple's CEO == Pixar's CEO.
Finally, for what it's worth, I'm a Mac user and a big OSX fan. But I know what my dollars are paying for and it ain't CPU cycles.
Regardless of whether the G5 is the fastest CPU for RenderMan, it is not per-CPU performance that matters. If you're setting up a rendering farm, you're buying n computers to render m frames per hour. At the end of the day, what matters is minimising $$$$$$ per m, not n, and I'll bet dollars to doughtnuts that commondity Intel/AMD whips a G5 mac in terms of rendered frames per dollar. Remember, Apple's CEO == Pixar's CEO.
Not if apple gives them to you below cost as a 'marketing expense'
autopr0n is like, down and stuff.
Pixar is migrating to OS X, primarily because of the G5. Pixar's OS/machine of choice seems to vary with the wind, whatever is the most powerful at the time. It seems that they believe the G5 is where it's at in the near future.
They're posting jobs for techs to assist in a migration to OS X.
Pixar hasn't used Macs every for primary work to my knowledge and Steve Jobs has been CEO there for YEARS. He's smart enough to let them pick the best horsepower to do the job.
Steve's been back at Apple since 1997 (nearly 6 years), and hasn't mandated a switch to Macs. If he did, you'd have heard an anonymous outpouring of complaints. But what you hear is, the G5 smokes and that they're migrating to Macs. This looks like the people doing the work made the decision.
Now, that isn't to say that in a year or 2 they don't switch to Itanium 2s or Opterons. I'm sure Pixar will continue to choose the most powerful machines for their type of work as they have done so in the past (SGI, Sun, Linux, Apple, etc)...